First-year engineering student at MIT-WPU Pune with a focus on embedded systems, computer architecture, and machine learning on constrained hardware. I build things that have to work outside controlled environments.
Currently applying for the LFX Mentorship Summer 2026 program under Kurt Keville (MIT) — porting Community AI/ML and HPC applications to RISC-V.
🥇 Gold Medal — First Place, Sci Quest 2025, MIT-WPU Built a TensorFlow/Keras ensemble CNN for groundnut leaf disease detection across 6 classes. Trained, optimised for inference efficiency, and deployed on Hugging Face. Competed against students from all branches and years of the institute.
- RISC-V HPC Porting — cross-compiling numerical methods (Bisection, RK4, LU Decomposition) for RV64GC, verified under QEMU
- Fixed-Point Arithmetic Library — portable C library for RISC-V cores without an FPU
- TF Lite on RISC-V — porting ML inference to RISC-V as part of LFX 2026 preparation
| Project | Description | Stack |
|---|---|---|
| riscv-hpc-port | Numerical methods cross-compiled for RV64GC, verified under QEMU | C, GCC, QEMU |
| Pretrained-CNN | Gold Medal Sci Quest 2025 — groundnut leaf disease detection | TensorFlow, Keras, Streamlit |
| fixed-point-riscv | Portable fixed-point arithmetic for RISC-V without FPU | C |
| rv32i-decoder | CLI tool decoding 32-bit RISC-V machine words to assembly | Python |
MIT World Peace University, Pune · tanmaygulhane12@gmail.com